Xenopus laevis cement gland as an experimental model for embryonic differentiation: II. The competence of embryonic cells

Ectoblastic cells of young Xenopus laevis gastrulae explanted in standard salt solutions for 5 days produce atypical epidermis. However, when these explants are first incubated for 6 h in Holtfreter solution containing 10 mM ammonium chloride at pH 7·5 and are then transferred for 5 days to a standard salt solution they undergo differentiation into typical cement gland (Picard, 1975). Under these in vitro conditions about 80 to 90 % of the explanted tissue is cement gland and the additional 20 to 10 % remains undifferentiated.

The aim of the present article is to specify more precisely which cells of the embryo are able to differentiate into cement gland after stimulation with 10 mM ammonium chloride. This problem deals with the position of the cells in the embryo and the time they become competent to respond to the stimulation. We also wanted to know whether the fate of gastrula cells which differentiate in vivo and in vitro into ectodermal or mesodermal tissues may be changed by ammonium chloride stimulation and forced to differentiate into cement gland. Finally, the variability of the results according to the egg-batch will be estimated.

The animals (Xenopus laevis, Daudin), the methods and the experimental conditions were the same as previously described (Picard, 1975). The volumes of individual tissues were estimated by morphometry of histological sections (Hennig & Meyer-Arendi, 1963).

Explants were prepared from embryos ranging from stage 8 to stage (Nieuwkoop & Faber, 1967). The explanted tissue was a cup of the blastocoelic roof extending from about 20° below the animal pole dorsally to about 70° below the animal pole ventrally.

The explants were first incubated for 6 h in a modified Holtfreter solution containing 10 mM ammonium chloride and 10 mM sodium bicarbonate at pH 7·6. After this stimulation they were cultured for 5 days in a standard salt solution (Barth & Barth, 1959), processed for histological examination and the individual volumes estimated.

The mean total volume of individual explants is given in Fig. 1 A. This volume decreased progressively to reach at stage a value of 59 % of the volume observed when the explants were prepared from embryos at stage 8. The total volume of explants from embryos at stages 11 and was slightly higher than the volume observed at stage ⁠.

The mean volume of cement gland per explant is given for each stage in Fig. 1B. The response of the ectoderm to incubation in ammonium chloride was weak at stages 8 and 9. The best differentiation occurred in stage 10 explants. Explants from older gastrulae formed smaller cement glands under the same conditions.

The percentage of the total explanted tissue occupied by the cement gland was low at stages 8 and 9, reached a maximum (62 %) at stages 10 and ⁠, and finally decreased to a value of about 2 % at stage (Fig. 1C).

The cement gland was the only differentiated tissue to be observed in the explants. The remaining tissue was atypical epidermis. Therefore, in each series the total explanted tissue is the sum of the undifferentiated tissue and the cement gland tissue. This undifferentiated tissue was predominant in stage-8 explants, decreased at stage to about 24 % of the stage-8 value and was again predominant at stage (Fig. 1D).

Two series of circular pieces of the ectoderm were dissected from stage-10 embryos. The centre of the explants was situated in the sagittal plane. The explants extended from 5° to 50° from the animal pole, ventrally in the first series and dorsally in the second one (Fig. 2).

The explants were stimulated for 6 h in 10 mM ammonium chloride as described previously. The volumes of the tissues were measured after 5 days of culture in Barth’s solution.

Table 1 shows that the ventral and dorsal portions of the ectoblast have about the same ability to respond to the stimulation by producing cement gland tissue. The mean cement gland volumes per explant are not statistically different. The percentages of the total explanted tissue occupied by the cement gland are very similar in both the ventral and dorsal portions of the ectoblast. No other differentiated tissue was observed in these explants. The mean total explanted tissue was significantly lower in the series of dorsal ectoderm than in the series of ventral ectoderm. This may be explained either by a more important tissue loss in the dorsal explants during the incubation or more probably by a smaller initial volume due to restraint during the dissection of dorsal explants to avoid reaching mesoblastic tissues.

The upper blastoporal lip of stage-10 embryos was dissected. The lower limit of the explants was the rim of the lip and the upper limit was about 5° above the equatorial plane (Fig. 2). Laterally, the explants extended to about 15° from the plane of symmetry.

The explanted lips were divided into two series. The control series was first cultured for 6 h in Holtfreter solution and then transferred for 5 days in Barth’s solution. The experimental series was incubated for 6 h in Holtfreter solution containing 10 mM ammonium chloride and 10 mM sodium bicarbonate. After this incubation the explants were cultured for 5 days in Barth’s solution.

The volumes of the tissues are given in Table 2. Cement gland, neural tissue and notochord were measured separately. The volumes of the remaining tissues, containing muscle, kidney tubules and endodermal structures, were measured together. The mean total volume of the explants remaining after 5 days of culture was markedly lower in the experimental series than in the controls. The volume of the former was about 59 % that of the latter. The difference is highly significant. The observation of living explants had shown that they were losing cells during the first 3 days of culture mostly from the underlying layer.

All the tissues which were measured had a volume significantly smaller in the experimental series than in the control series. The notochord was the most affected and accounted for only 28·6 % of the volume observed in the controls. This proportion was respectively 47·7 % and 60·4 % for cement gland and neural tissue.

The percentage of the total explanted tissue occupied by cement gland and neural tissue was approximately equal in the two culture series. On the other hand this percentage was only 7·7 % for notochordal tissue in the experimental series compared with 16·0 % in the control series.

In order to estimate the reproducibility of the results, ectodermal explants from stage-10 embryos belonging to twelve different egg-batches were stimulated for 6 h by 10 mM ammonium chloride and subsequently cultured for 5 days in Barth’s solution (Table 3).

The mean volume of the cement gland per explant varied from 10·3 × 10⁶ μm³ to 36·0 × 10⁶ μm³. The weighted mean was 24·9 × 10⁶ μm³. The percentage of the total explanted tissue occupied by the cement gland varied from 26· 1 % to 80·4 % and the percentage corresponding to the weighted means was 55·0 %.

The weighted means of the total and of the undifferentiated tissue per explant were 45·1 × 10⁶/mi³ and 20·3 × 10⁶μm³, respectively.

In order to check the homogeneity of the variances Bartlett’s test was applied to these data (Snedecor, 1959). Calculations indicated a ; χ² of 62·8 (P ⪡0·01) for undifferentiated tissue, of 14·0 (P < 0·3) for cement gland tissue and of 26·7 (P < 0·01) for total explanted tissue. The variances are therefore only homogeneous for the volumes of the cement gland. The heterogeneity of variances of the total explanted tissue appears to be essentially the result of the heterogeneity observed in the undifferentiated tissue.

Analysis of variance was applied to the data corresponding to the volumes of the cement gland and the variance ratio was calculated to be F = 14·0 (P ⪡ 0·01). Hence the amount of cement gland appearing in the explants as a result of a 6 h ammonium chloride stimulation varies significantly according to different egg-batches.

The response of ectoblastic cells to a 6 h stimulation with ammonium chloride varies considerably according to the stage of the embryos. The beginning of gastrulation (stage 10) is the time when the explants produce the greatest amount of cement gland tissue under these conditions (Fig. 1B). This stage may be accepted as the time of optimum competence and will be routinely used in further studies on this experimental mode. The mean volume of the total tissue per explant measured after 5 days culture in Barth’s solution also varied according to the stage of the embryos (Fig. 1 A). The lowest value was observed at stage 10. This may probably be explained by the epibolic movements occurring during gastrulation. Indeed, as a result of these cellular movements the blastocoelic roof becomes thinner and hence the same explanted area will have a smaller starting volume as gastrulation proceeds. A slight increase of the mean volume of the total tissue per explant occurs at stages 11 and ⁠. This probably results from a better resistance of older ectoblastic cells to incubation in ammonium chloride. Indeed, we observed that explants from stage 11 and lose fewer dissociated cells than younger explants.

In these experiments a large portion of the blastocoelic roof was explanted. If smaller explants had been used, the results could have been very different according to the presence or absence of the presumptive cement gland area in the explanted tissue. It could indeed be possible that only this presumptive area is able to respond to ammonium chloride stimulation and that the remaining ectoblast would only produce atypical epidermis. The presumptive area of the cement gland on amphibian gastrulae is located either ventrally (Discoglossus) or dorsally (Bombinator) as regards the animal pole of the embryo (Vogt, 1929). The fate map of Xenopus laevis gastrula is not known. However, the fate maps of the two above-mentioned anuran species indicate that the presumptive areas of their cement glands do not exceed 10° in the sagittal plane. This most probably applies to the presumptive area in Xenopus laevis too. Therefore, in the experiment described in section 2 the presumptive area of the cement gland will not be situated in both explanted portions. The volume of the cement gland was not significantly different in these two portions (Table 1). It is therefore safe to conclude that other parts of the ectoblast and possibly the whole ectoblast is able to differentiate into cement gland in response to ammonium chloride stimulation. Hence the fate of certain parts of the ectoblast can be changed by this treatment. Moreover, if the presumptive area was situated in one of the two explanted portions it would appear that the ability of that area to produce cement gland tissue in response to ammonium chloride stimulation is not greater than that of the remaining ectoblast.

It is well known that blastoporal lips of young amphibian gastrulae explanted in a physiological salt solution will differentiate mainly into cephalic endodermal and mesodermal structures (Holtfreter, 1938). When Xenopus laevis blastoporal lips were incubated for the first 6 h in 10 mM ammonium chloride the treatment was unable to change massively the fate of this tissue (Table 2). Under these conditions the volume of the total tissue of explanted blastoporal lips remaining after 5 days culture was 41 % lower than in the controls. All the tissues suffered from the loss of material and the most affected one was the notochord. The volume of the cement gland was also smaller in the experimental series. This is probably not due to a direct inhibition of the differentiation into cement gland. Indeed, the explants contain the natural inductors of the cement gland, i.e. endoderm and/or brain (Yamada, 1933, 1938). As the latter tissues are affected by the treatment, it seems more likely that the smaller volume of the cement gland is due to an indirect influence.

The response of the explants to incubation in ammonium chloride displays some quantitative variations according to egg-batches. The 12 identical experiments performed over a five-month period are insufficient to disclose a relationship with seasonal changes. The most important variability was observed in the undifferentiated tissue. The volume of the cement gland appearing in the explants also differs significantly according to the egg-batches. This biological variability is not surprising and may have numerous causes which are impossible to analyse.

In these series the weighted mean of the volume of the cement gland per explant was 24·9 × 10⁶ μm³. This accounts for more than twice the volume of an individual cement gland differentiated in vivo (unpublished results). Cells of cement glands differentiated in vivo and in vitro will probably display approximately the same volume. Therefore these results represent an independent confirmation that the ability of the ectoblast to differentiate into cement gland in response of ammonium chloride stimulation is found in a broader region than the presumptive area.

On average the cement gland tissue accounted for 55 % of the total tissue measured after 5 days of culture. The cement gland originates exclusively from the superficial layer of the ectoderm (Lieberkind, 1937). After 5 days of culture it is not possible to detect by histological means what part of the undifferentiated tissue originates either from the superficial or from the underlying layer of the ectoderm. However, results to be published elsewhere have demonstrated that the volume of the underlying layer accounts for about 62 % of the ectodermal portion at the time of explantation and that the greater part of this tissue remains in the explants during cultivation. Therefore most of the undifferentiated tissue (45 % of total tissue) originates from the underlying ectodermal layer. A rough estimation indicates that after a 5-day culture the cement gland accounts for 80–100 % of the volume of the superficial layer.

This conclusion is interesting in view of the possibility of isolating cells from the latter layer by density gradient centrifugation (Morrill & Kostellow, 1965; MacMurdo & Zalik, 1970, 1971). This improvement would reduce further the volume of contaminating undifferentiated tissue.

Finally, the absence of any significant amount of any other differentiated tissue in the stimulated explants is important because the homogeneity of the material will facilitate interpretation of biochemical studies.